Endosymbionts Reduce Microbiome Diversity and Modify Host Metabolism and Fecundity in the Planthopper Sogatella furcifera

Abstract

Endosymbionts can strongly affect bacterial microbiota in pests. The white-backed planthopper Sogatella furcifera, a notorious pest in rice, is usually co-infected with Cardinium and Wolbachia, but the effects of these endosymbionts together or individually on the host microbiome and fecundity are unclear. Here, we established three S. furcifera lines (Cardinium and Wolbachia double-infected, Cardinium single-infected, and both-uninfected lines) backcrossed to a common nuclear background and found that single and double infections reduced bacterial diversity and changed bacterial community structure across nymph and adult stages and across adult tissues. The endosymbionts differed in densities between adults and nymphs as well as across adult tissues, with the distribution of Cardinium affected by Wolbachia. Both the single infection and particularly the double infection reduced host fecundity. Lines also differed in levels of metabolites, some of which may influence fecundity (e.g., arginine biosynthesis and nicotinamide metabolism). Cardinium in the single-infected line upregulated metabolic levels, while Wolbachia in the double-infected line appeared to mainly downregulate them. Association analysis pointed to possible connections between various bacteria and differential metabolites. These results reveal that Cardinium by itself and in combination with Wolbachia affect bacterial microbiota and levels of metabolites, with likely effects on host fecundity. Many of the effects of these metabolically limited endosymbionts that are dependent on the hosts may be exerted through manipulation of the microbiome. IMPORTANCE Endosymbionts can profoundly affect the nutrition, immunity, development, and reproduction of insect hosts, but the effects of multiple endosymbiont infections on microbiota and the interaction of these effects with insect host fitness are not well known. By establishing S. furcifera lines with different endosymbiont infection status, we found that Cardinium and the combined Cardinium + Wolbachia infections differentially reduced bacterial diversity as well as changing bacterial community structure and affecting metabolism, which may connect to negative fitness effects of the endosymbionts on their host. These results established the connections between reduced bacterial diversity, decreased fecundity and metabolic responses in S. furcifera.

Keywords: Cardinium and Wolbachia; Sogatella furcifera; bacterial microbiota; correlation; metabolite; microbiome; tissue-specific.

FIG 1

Bacterial diversity and relative abundance across different developmental stages in the three S. furcifera lines. (A) Principal coordinates analysis (PCoA) showing bacterial OTU-based clusters. Circles with different colors represent different samples, n = 3. (B and C) Scatterplots representing Chao1 (B) and Fisher (C) indexes of bacterial diversity. Asterisks indicate significant differences between lines, n = 3; *, P < 0.05; **, P < 0.01; ***, P < 0.001. (D) Relative abundance of bacteria showing bacterial community structure at genus level. Bacteria with a relative abundance of less than 1% were classified as Others. U, uninfected S. furcifera line; C, Cardinium single-infected S. furcifera line; CW, Cardinium and Wolbachia double-infected S. furcifera line; N, nymph; F, female; M, male.

FIG 2

Correlation analysis of bacteria and host fecundity determination in the three S. furcifera lines. (A–C) Pairwise co-occurrence patterns of bacterial microbiota at phylum-level in U (A), C (B) and CW (C). Red, blue and gray indicate positive correlation, negative correlation and no correlation between the two bacterial phyla, respectively. (D and E) Genus-level correlation of Cardinium (D) and Wolbachia (E) with other bacteria in double-infected line (CW). (F) Fecundity effects of the three lines. Asterisks indicate significant difference the two compared group, *, P < 0.05; **, P < 0.01; ***, P < 0.001. Data were expressed as mean ± SD. U, uninfected S. furcifera line; C, Cardinium single-infected S. furcifera line; CW, Cardinium and Wolbachia double-infected S. furcifera line.

FIG 3

Densities and distributions of bacteria and endosymbionts in three S. furcifera lines. (A–F) Box plots representing the relative densities of total bacteria, endosymbionts (Cardinium and Wolbachia) and Acinetobacter across different developmental stages (A–C) and in adult tissues (D–F) of the three lines. Asterisks indicate significant difference between the infected group (C or CW) and uninfected group (U), n = 12; *, P < 0.05; **, P < 0.01; ***, P < 0.001. (G–I) Distributions of Cardinium, Wolbachia and nuclear DNA in the different adult tissues of the three lines. Ovariole (G), testis (H) and midgut (I). Bacteria, Cardinium and Wolbachia were stained as green, red and fuchsia, respectively, using specific probes. Corresponding arrows indicate the staining regions. Scale bars are 100 μm. U, uninfected S. furcifera line; C, Cardinium single-infected S. furcifera line; CW, Cardinium and Wolbachia double-infected S. furcifera line.

FIG 4

Metabolite analysis for the three S. furcifera lines. (A) The number of total differential metabolites and identified upregulated and downregulated differential metabolites for each comparison in positive ion mode and negative ion mode. (B) Principal Component Analysis (PCA) showing the sample clusters for the three lines. Coordinate axis represents the percent contribution of each principal component to the total variance. (C) Venn diagrams depicting the number of differentially expressed genes (DEGs) in the three comparisons. (D–G) Bubble diagrams indicating the Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment of differential metabolites. The differential metabolites include upregulated differential metabolites in the C/U comparison (D), upregulated (E) and downregulated (F) differential metabolites in the CW/U comparison and downregulated differential metabolites in the CW/C comparison (G). U, uninfected S. furcifera line; C, Cardinium single-infected S. furcifera line; CW, Cardinium and Wolbachia double-infected S. furcifera line.

FIG 5

Correlation analysis of differential microbes and metabolites in the three group comparisons. Pairwise differential microbe-metabolite correlation analysis based on Spearman correlation analysis in C/U (A), CW/U (B) and CW/C (C) comparisons. Microbiome and metabolome data are shown only when coefficient R-value was > 0.9 or < −0.9 and P value was < 0.05. Red and green lines indicate positive and negative correlations, respectively. U, uninfected S. furcifera line; C, Cardinium single-infected S. furcifera line; CW, Cardinium and Wolbachia double-infected S. furcifera line.

FIG 6

Model for the potential correlations between bacterial diversity, metabolic response and host fecundity under the infection of endosymbionts. The differences in endosymbionts, bacterial diversity, metabolites, host fecundity and their potential correlations were described. The number of bacteria represents the level of bacterial diversity. The representative differential compounds in the three lines are described. Orange and blue, respectively, indicate the compounds with relatively high and low concentrations in their comparisons between the two lines. The compound names were marked below the structural formula, and the names of the two lines being compared were marked behind the name. The fecundity effects of the three lines were expressed by the number of eggs. Orange and blue texts next to the arrows, respectively, indicate the upregulated and downregulated KEGG metabolic pathways at the arrow directions.